Ascites, Neoadjuvant Chemotherapy, and Ovarian Cancer

Abstract & Commentary 

Synopsis: This study performed a retrospective analysis of ovarian cancer patients and found that residual tumor and volume of ascites were independent prognostic factors. They then performed a prospective, nonrandomized phase II study comparing primary surgery with neoadjuvant chemotherapy in patients with large volume ascites (> 500 cc). Their results suggest an improved survival rate if neoadjuvant chemotherapy is given to these patients.

Source:Kuhn W, et al.Cancer. 2001;92:2585-2591.

From 1996 to 2000, 31 patients with laparoscopically diagnosed stage IIIC ovarian cancer with large volume ascites (> 500 cc) were enrolled in a study of neoadjuvant chemotherapy. Treatment consisted of 3 preoperative cycles of carboplatin (AUC = 5) and paclitaxel (175 mg/m2), and 3 additional cycles postoperatively. Those considered to be "in poor general health" had the paclitaxel omitted.

A control group consisted of 32 comparable patients treated with surgery followed by 6 cycles of chemotherapy during the same time period. These patients had either refused entry onto the protocol or did not appear "compatible for psychological reasons." In this prospective, but nonrandomized, analysis, Kuhn and colleagues found that neoadjuvant therapy improved the median survival for patients with large volume ascites (42 vs 23 months, P = 0.0007).

Comment by Kenneth W. Kotz, MD

Over the years, as survival in ovarian cancer was highly dependent on optimal cytoreduction, the standard of care has been aggressive surgical debulking followed by chemotherapy. The theoretical advantage of this approach includes improved chemosensitivity because large necrotic tumors with marginal blood supply have been removed leaving behind the smaller residual tumors with a higher growth fraction.

Patients with large volume ascites are less amenable to optimal primary cytoreduction and may have a worse prognosis. A retrospective, multivariate analysis was performed on 193 stage IIIC ovarian cancer patients treated between 1982 and 1995. Kuhn et al found that postoperative residual tumor (> 2 cm) had the strongest effect on prognosis, but that the volume of ascites was also an independent prognostic factor. For example, patients with > 500 cc ascites had a median survival of 19 months compared to 53 months for patients with < 500 cc.

The subsequent prospective analysis of these higher-risk patients with large-volume ascites showed that suboptimal debulking (> 2 cm residual) occurred less often in patients treated with neoadjuvant chemotherapy (16% vs 37%; P = 0.04). Of greater importance, however, is that the neoadjuvant chemotherapy was associated with a survival advantage (median survival, 42 vs 23 months). No difference in surgical morbidity or mortality emerged.

Even though the 2 prospectively studied groups were well balanced with respect to age, histology, grade, and lymph node status, bias may have been introduced by the lack of randomization. Specifically, the exclusion of patients not appearing psychologically fit might have shifted sicker patients into the control group. In addition, Kuhn et al point out that a bias toward more aggressive debulking could have occurred in patients treated with preoperative chemotherapy. Whether the 2 groups were balanced with respect to the omission of paclitaxel and delivery of planned chemotherapy was not reported.

Even in the platinum/paclitaxel era, the proportion of patients with stage III/IV disease who undergo maximal cytoreduction remains one of the most important predictors of survival.1 Interval cytoreductive surgery has also been shown to improve survival for patients in whom the primary surgery resulted in suboptimally debulked disease.2,3 Therefore, the improved survival from neoadjuvant therapy noted by Kuhn et al may be the result of improved resectability after downstaging. In fact, a multivariate analysis by Kuhn et al showed that residual tumor, but not timing of chemotherapy, had a statistically significant prognostic value. Unfortunately, Kuhn et al did not report response rates or results of serum tumor markers.

Neoadjuvant therapy may also be appropriate to reduce surgical morbidity. Lower postoperative mortality rates,4 as well as shorter surgical time and fewer transfusions,5,6 have been reported with neoadjuvant therapy. Higher-risk patients in whom surgery might be delayed in order to institute chemotherapy include those with uncountable plaques, tumor load > 1 gm, a poor performance status,4 poor nutritional status, or stage IV disease (excluding pleural effusion).4,6 Those not responding to chemotherapy could be identified and approached more palliatively.

To summarize, neoadjuvant chemotherapy may be useful to reduce surgical morbidity in certain high-risk patients. In patients with large-volume ascites, the results of Kuhn et al suggest, but do not prove, that there may even be a survival advantage. The applicability of this approach to other patients with ovarian cancer is unknown, but does not seem to be associated with a worse outcome.4,6 


1. Bristow R, et al. Proceedings of the American Society of Clinical Oncologists 2001; Poster Session #807.

2. Ansquer Y et al. Cancer. 2001;91:2329-2334.

3. van der Burg M, et al. N Engl J Med. 1995;332: 629-634.

4. Vergote I, et al. Semin Oncol. 2000;27:31-36.

5. Surwit E et al. Int J Gynecol Cancer. 1996;6:356-361.

6. Schwartz P, et al. Gynecol Oncol. 1999;72:93-99.